Wrong code detection is it a firmware problem?

[FMB]

Advertisement (gone after registration)

My new M9, since the day I got it, has code detection problems. Reads Cron 50 instead Lux 50, Cron 35 and Tri-Elmar as uncoded , 24 Elmarit and Cron 90 reads OK and so and so.

 

Do you think is there a problem of hardware or firmware? I've just re-installed v. 1002 and nothing has changed, the wrong detection is confirmed.

 

 

Francisco

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[vanhulsenbeek]

Have you checked if the little"window" on the camera that detects the codes is clean? Are the codefields on your lenses clean? If so, most probably a hardware problem indeed.

[FMB]

Sander, my answer your two questions isYES. I'm afraid you are right!

 

Francisco

[jaapv]

Have you checked the codes of the lenses against Carstens list?

[Shootist]

Also make sure the lenses that are not being read correctly are properly locked into the mount. IE turn enough to lock proprerly and not overturned to go by the lock.

 

I do this by never touching the lock release button when installing a lens and listening for the click the lock make when it seats into the lens.

[danyves]

Check your profiles.

[scott kirkpatrick]

Advertisement (gone after registration)

For eachmisidentification that you have seen, write out the 6-bit code that is on the lens, which is presumably correct. Then just below it, write in the 6-bit code for the lens that your camera thinks it is. If the codes always differ in the same bit position or positions, the problem is in your diode detecters. Hardware; trip to the factory for warranty service, but at least you know.

 

scott

[FMB]

Jaap, Shootist,danyves and Scot, I've forgotten to say a very important test I've done: with my M8 camera all the lenses have been correctly identified.

 

Sorry, but what I wanted to know was if there were a system, like you have proposed Scot, to check better the camera or if the issue could be a firmware problem.

 

Thank you all.

 

Francisco

[Shootist]

Jaap, Shootist,danyves and Scot, I've forgotten to say a very important test I've done: with my M8 camera all the lenses have been correctly identified.

 

Sorry, but what I wanted to know was if there were a system, like you have proposed Scot, to check better the camera or if the issue could be a firmware problem.

 

Thank you all.

 

Francisco

 

It certainly is not a firmware problem.

[ho_co]

Francisco:

Scott's method is clear and simple.

 

Mark Norton demonstrated a different method some years back that you could easily use with the M8 or any other IR-sensitive camera:

 

Go into a dark space. Turn on the M9 without lens. Using the IR-sensitive camera (without IR-Cut filter, of course), photograph the front of the M9. That will show you which of the lens-ID diodes are working and which isn't.

 

Mark's method is fun and photographic. Scott's and Mark's will both produce the same incontrovertible evidence.

[marknorton]

You beat me to it, HC!

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[ho_co]

No, Mark!

 

I spent half an hour looking for that shot because it so perfectly defines the M8. Your thread introduced it with wonderful irony and humor.

 

Is the thread still available? What's the link?

 

 

(By "so perfectly defines the M8," I refer to the camera's use of IR to achieve lens recognition, and its opposing need to filter it out to make photos. )

[Bo_Lorentzen]

Just a side note, I have a couple of hand coded lenses, when I got the M9 these did not work correctly until I repainted the black and cleaned the white areas, it seems like the M9 is more sensitive to the quality of the black.

 

Naturally this observation do not apply to Leica coded lenses.

 

.

[scott kirkpatrick]

Mark's test tells you if the LED's in the six little windows are lighting up. But the second step in lens recognition is that there is a receiver next to each LED . The receiver has to see the light reflected off a white spot (or not see the light absorbed in a dark spot) in order to register the six bit pattern. If the receiver is stuck ON or OFF, that will cause problems in lens identification as well.

 

scott

[adan]

Just for the record, I recently had one coded lens (21mm) go MIA on the M9 for a run of about 7 shots. It is back to normal now.

 

It could have been either 1) slightly sloppy mounting that didn't click in all the way (a quick lens change during a news shoot), or 2) grime in the white spots - I flicked out a little bit of "something" about .3mm square that had settled into one of the white indents.

 

Didn't affect the pix any - I only noticed because I was checking the EXIF to see if those shots had been made with the 35 or 21.

[scott kirkpatrick]

OK, OK. I know things with bits in them are a little off-putting, but really this isn't hard. I looked at Carsten's tables and checked the OP's first observation. The Lux 50 is interpreted as a Cron 50. Well the bit patterns are:

 

LUX50 1 0 0 0 0 0

CRON50 1 0 0 0 0 1

 

Hmm, maybe the receiver for the last bit is stuck ON?

 

Let's look at the CRON 35 0 1 1 1 0 . Is there a lens with pattern 0 1 1 1 1 1 which it could be confused with? Yes, the 0.95 Nocitilux, but the frame lever would be in the wrong position, so the M9 would reject that lens ID and call it not-recognized.

 

Now MATE 1 0 1 0 1 0 would be read as 1 0 1 0 1 1 if this theory is right. That's a Summarit 35. Which would expect to see the frame lever in the 35 position. Is the MATE recognized when it is set to 35 mm, but not when it is set to the other focal lengths? Or if you meant the WATE 0 1 0 0 0 0 , that would be read as a 0 1 0 0 0 1 , and I couldn't find such a lens in Carsten's list.

 

The 24 Elmarit-asph is coded 0 1 1 0 0 1 and the Croin 90 APO is 1 0 0 1 0 1, so these are read correctly.

 

So it certainly sounds like a very specific hardware problem involving the devices which read the last bit. Two further tests (put on the cron 35 and hold the frame lever in the 50 position while taking a shot; put on the MATE -- if that's what you have -- and set it as a 35) will add supporting evidence. It seems a shame that a $7000 camera could misbehave because of one component almost too small to see with the naked eye, but that's surface mount LSI.

 

I think you are owed a camera-swap, and someone will get a refurb with the little card that holds the six diodes replaced. But I wouldn't give up an M9 in the hand for a month just to get this little hassle fixed. See what you can arrange for a loaner or something.

 

scott

[FMB]

Scott, yesterday night following your method I arrived just to the same conclusion and that means that my M9 must to travel to Germany and I'll be a lot of days "only" with my infallible M8.

 

I want to try your method, Mark, but I have not macro lens and may be the test could fail.

Nevertheless I trust in my M8!

 

Thank you all people participating with your comments.

 

Francisco

[scott kirkpatrick]

If you have a 90 Cron and your M8, just set it to its closest focusing distance (0.7 m, isn't it?), remove the IR filter, and shoot the front of your M9 with the diodes exposed. I suppose it will work best if you set this up with a tripod, focus carefully and only then turn the lights off. You'll probably have to use trial and error with manual exposure to get the six (or five) little bright lights to show up. Blowing up the image to see what the red lights look like shouldn't be a problem.

 

I just realized that the six bit convention is that a black dot (no light) is a "1" so you might indeed see that LED number 6 is dark. That will eliminate any doubts about mounting errors, etc.

 

scott

[Advertisement]

Hallo,

Look here: Leica M

[marknorton]

Couple of further thoughts.

 

First, Scott is entirely correct that the fact the LEDs are working doesn't mean the photodiodes and the transistors which amplify the signal back from them are.

 

Second, The M9 may be different from the M8 which was looking for a lens any time the camera was powered so that the focal length selection menu for the WATE could pop up as soon as one is mounted. That may not be true of the M9.

 

[For new readers: whereas the original Tri-Elmar ("MATE") coupled the selected focal length to the camera to select the required framelines, this is not true of the 16-18-21 Tri-Elmar ("WATE") which is wider than any of the built-in framelines. However, Leica discovered too late that the camera needed to know the selected focal length anyway for IR filter vignetting correction which is why there is a manual selection required when using the WATE].

 

Third, if the LEDs, photo-diodes or the other circuitry have failed, it's a major job to replace; the entire camera has to come apart aside from the rangefinder. On that basis, I would press for a new camera, not one which has had to have such major surgery.

 

This is how is in the M8, likely identical in the M9:

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

 

[marknorton]

No, Mark!

 

I spent half an hour looking for that shot because it so perfectly defines the M8. Your thread introduced it with wonderful irony and humor.

 

Is the thread still available? What's the link?

 

 

(By "so perfectly defines the M8," I refer to the camera's use of IR to achieve lens recognition, and its opposing need to filter it out to make photos. )

 

HC, thanks, here's the original thread...

 

http://www.l-camera-forum.com/leica-forum/leica-m8-forum/23619-lens-sensor-action.html#post247347

 

I remember my frustration not being able to capture the LEDs first with a Nikon (not IR sensitive) and then with another M8, only to realise, (hand smacks forehead), I needed to remove the IR filter.